High-resolution building stock energy model for assessing the impact on the load curve of residential space heating electrification

Assessing the impact of space heating electrification on the power demand requires high-resolution building stock energy models. A bottom-up white-box model offers the possibility to explicitly account for the impact of energy efficiency measures or behavioral changes. Although, these models require significant parametrization effort, and while they can reach the appropriate level of resolution, their validation is often performed at a coarse resolution.

This paper proposes a calibration and validation of the hourly space heating electrical load curve of the French residential stock. A representative sample of the French households based on a survey of 4,000 households is used to simulate the residential space heating load curve with simplified lumped capacitance (RC) models. A representative panel of the French household consisting of 6,000 smart meters is used to build a statistical estimation of the residential load curve.

At the national level, the model is accurate with a (CV)RMSE of 9% and a NMBE of 3%. The model remains quite accurate at finer scales. For 25 specific segments of the stock, the median (CV)RMSE is 22%, NMBE is 10% even if a few segments exhibit higher deviation or compensation errors. The model is fit for assessing future load curve warping based on explicit changes in energy efficiency of sufficiency measures.